期刊
2018 55TH ACM/ESDA/IEEE DESIGN AUTOMATION CONFERENCE (DAC)
卷 -, 期 -, 页码 -出版社
IEEE
DOI: 10.1145/3195970.3195998
关键词
-
资金
- NSF [CCF-1744082, CNS-1744111]
- DOE [SC0017030]
Although ReRAM-based convolutional neural network (CNN) accelerators have been widely studied, state-of-the-art solutions suffer from either incapability of training (e.g., ISSAC [1]) or inefficiency of inference (e.g., PipeLayer [2]) due to the pipeline design. In this work, we propose AtomLayer-a universal ReRAM-based accelerator to support both efficient CNN training and inference. AtomLayer uses the atomic layer computation which processes only one network layer each time to eliminate the pipeline related issues such as long latency, pipeline bubbles and large on-chip buffer overhead. For further optimization, we use a unique filter mapping and a data reuse system to minimize the cost of layer switching and DRAM access. Our experimental results show that AtomLayer can achieve higher power efficiency than ISSAC in inference (1.1x) and PipeLayer in training (1.6x), respectively, meanwhile reducing the footprint by 15x.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据